|Publication number||US6975325 B2|
|Application number||US 10/763,782|
|Publication date||Dec 13, 2005|
|Filing date||Jan 23, 2004|
|Priority date||Jan 23, 2004|
|Also published as||US20050162437|
|Publication number||10763782, 763782, US 6975325 B2, US 6975325B2, US-B2-6975325, US6975325 B2, US6975325B2|
|Inventors||Stephen L. Morein, Tom E. Frisinger, Philip J. Rogers, Richard Bagley|
|Original Assignee||Ati Technologies Inc.|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (4), Referenced by (8), Classifications (8), Legal Events (5)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The invention relates generally to graphics processing and more particularly to the utilization of state and shader management in graphics processing.
When processing graphics for visual display, prior to rendering the graphics, the system receives and must compile new vector inputs. This process can be a time consuming step, due to the amount of information which must be compiled. Once all of the input vector information is compiled, a hardware state and shader vector is created and then provided to a graphics processor. The graphics processor utilizes the state and shader vector to produce a visual output.
This graphics processing system can be extremely inefficient when a large input vector needs to be compiled and provided to the processor. Furthermore, to increase optimization for the best performance of the graphics processor, the compiler may require a long compile time to properly compile the large input vector.
In conventional processing systems, it is not uncommon to have the compiler recompiling commonly occurring large input vectors. Recompilation of the same input vector is redundant and wasteful.
The compilation process illustrated in
The invention will be more readily understood with reference to the following drawings wherein:
The disclosed invention is an apparatus and method for graphics processing including providing an abstract state vector to a state and shader cache and determining if a cache entry exists within the state and shader cache for the abstract state vector, wherein the cache entry contains a hardware state and shader vector. If the cache entry exists, the hardware state and shader vector which is within the cache entry is provided to a graphics processor. Also, when the abstract state vector is provided to the state and shader cache, it is also provided to a compiler. Therefore, if the cache entry does not exist in the hardware state and shader vector, the compiler compiles the abstract state vector to thereupon produce the hardware state and shader vector. The state and shader cache generates a cache entry for the abstract state vector which was not contained therein and when the compiler has completed the compilation process, the hardware state and shader vector is written to the state and shader cache in the data entry having the abstract state vector. Thereupon, the state and shader cache provides the hardware state and shader vector to the graphics processor.
The abstract state vector is comprised of a vertex shader intermediate form, vertex shader constants, pixel shader intermediate form, pixel shader constants, and abstract render state. The cache entry within the state and shader cache is comprised of the abstract state vector and the hardware state and shader vector.
The abstract state vector 122 is provided to the state and shader cache 124 and the compiler 126 via connection 128. The state and shader cache 124 is further operably coupled with the compiler 126 via connection 130, wherein the state and shader cache 124 can provide the compiler 126 an indication when a miss has occurred. Furthermore, the compiler 126 is coupled to the state and shader cache 124 via connection 132, whereupon the compiler 126 can provide the compilation output to the state and shader cache 124. The state and shader cache 124 is further operably coupled to a hardware 134 via connection 136, wherein the state and shader cache 124 provides the hardware state and shader vector to the hardware 134. In one embodiment, the hardware may be a graphics processor 134 or any other suitable hardware device capable of recognizing or transferring the hardware state and shader vector.
Also illustrated in
The next step, 206, is determining if a cache entry 127 exists within the state and shader cache 124 for the abstract state vector. As illustrated in
Once the cache entry 127 has been completed, the hardware state and shader data within the cache entry 127 is provided to a graphics processor, such as the hardware 134 shown generally in
Thereupon, the graphics processor of
Moreover, it is also important for the state and shader cache to quickly and efficiently determine whether or not the abstract state vector has already been compiled and there is a cache entry containing the hardware state and shader vector. In one embodiment, a sub-vector approach is utilized to quickly and efficiently compare received abstract state vectors and compare them with abstract state vectors stored in the state and shader cache.
Concurrent with the abstract state sub-vector index 234, an array of the abstract state sub-vectors 236, wherein the abstract state sub-vector data is stored therein, is created. Each abstract state sub-vector index 234, looks up the corresponding abstract state sub-vector data 236 in that array. An array of state and shader hardware vectors 238 is created and data is added to this array each time the compiler is run. The combination of all indices 234 is used to index or address the array of state and shader hardware vectors 238. The combination of the indices 234 to create an addressing key may be a concatenation of the indices, a hash function or other method apparent to one of ordinary skill in the art. Thereupon, an array of the state and shader vectors is created wherein it is addressed by composite index, specifically the index A through H, 234.
If all of the abstract state sub-vectors are found, the next step is searching for a composite index in an array of composite indices and hardware state and shader vectors, step 246. Designated at 248 if the composite index is found, it is determined that a cache entry exists, designated at 250. If an abstract state sub-vector is not found, step 244 or the composite index is not found in the array of composite indices 248, it is determined that a cache entry does not exist, step 252. Thereupon the state and shader cache can make a determination whether to send a miss signal to the compiler, step 254.
It should be understood that the implementation of other variations and modifications of the invention and its various aspects will be apparent to those of ordinary skill in the art, and that the invention is not limited by the specific embodiments described herein. For example, the state and shader cache may contain more information beyond the abstract state vector, the hardware state and shader vector for each individual cache entry, as may be used by the hardware connected to the state and shader cache. It is therefore contemplated to cover by the present invention, any and all modifications, variations, or equivalents that fall within the spirit and scope of the basic underlying principles disclosed and claimed herein.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US5793374 *||Jul 28, 1995||Aug 11, 1998||Microsoft Corporation||Specialized shaders for shading objects in computer generated images|
|US6268875 *||Aug 4, 2000||Jul 31, 2001||Apple Computer, Inc.||Deferred shading graphics pipeline processor|
|US6339813 *||Jan 7, 2000||Jan 15, 2002||International Business Machines Corporation||Memory system for permitting simultaneous processor access to a cache line and sub-cache line sectors fill and writeback to a system memory|
|US20040207622 *||Mar 31, 2003||Oct 21, 2004||Deering Michael F.||Efficient implementation of shading language programs using controlled partial evaluation|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7825933 *||Feb 24, 2006||Nov 2, 2010||Nvidia Corporation||Managing primitive program vertex attributes as per-attribute arrays|
|US8006236||Feb 24, 2006||Aug 23, 2011||Nvidia Corporation||System and method for compiling high-level primitive programs into primitive program micro-code|
|US8171461||Feb 24, 2006||May 1, 2012||Nvidia Coporation||Primitive program compilation for flat attributes with provoking vertex independence|
|US8711159||Feb 23, 2009||Apr 29, 2014||Microsoft Corporation||VGPU: a real time GPU emulator|
|US8970608 *||Apr 1, 2011||Mar 3, 2015||Nvidia Corporation||State objects for specifying dynamic state|
|US20100188412 *||Jan 28, 2009||Jul 29, 2010||Microsoft Corporation||Content based cache for graphics resource management|
|US20100214301 *||Feb 23, 2009||Aug 26, 2010||Microsoft Corporation||VGPU: A real time GPU emulator|
|US20110242118 *||Apr 1, 2011||Oct 6, 2011||Bolz Jeffrey A||State Objects for Specifying Dynamic State|
|U.S. Classification||345/557, 345/559|
|International Classification||G09G5/36, G06T15/50|
|Cooperative Classification||G06T15/80, G06T15/50|
|European Classification||G06T15/50, G06T15/80|
|Sep 16, 2004||AS||Assignment|
Owner name: ATI TECHNOLOGIES, INC., ONTARIO
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:MOREIN, STEPHEN L.;FRISINGER, TOM E.;ROGERS, PHILIP J.;AND OTHERS;REEL/FRAME:015144/0117;SIGNING DATES FROM 20040120 TO 20040121
|May 21, 2009||FPAY||Fee payment|
Year of fee payment: 4
|May 12, 2011||AS||Assignment|
Owner name: ATI TECHNOLOGIES ULC, CANADA
Free format text: CHANGE OF NAME;ASSIGNOR:ATI TECHNOLOGIES INC.;REEL/FRAME:026270/0027
Effective date: 20061025
|Mar 18, 2013||FPAY||Fee payment|
Year of fee payment: 8
|Jun 1, 2017||FPAY||Fee payment|
Year of fee payment: 12